Shooting a monitor works better than you might think and I have gotten better results than many other methods and much faster. I have done this in the past when I was desperate.
I also recently achieved acceptable results downscaling in After Effects and compressing in Episode with some smoothing.
I just ordered a Cuda card as I have heard that downscaling with Cuda uses the much better Lanczos algorithm.
Often I wonder what they were thinking when they came up with HD, not to mention 160+ video formats. I just shot a show in DvcPro 50 SD 16:9 and it looks great. Most people think it is HD. It was shot with a Panny SDX 900. The images blow away my XDCam HD camera. This show looked better on a SD DVD than any hd material downscaled.
I’m going to try Eric’s advice first and see what happens. Michael, I have a Video card that I think it has Cuda cores (GeForce GTX 550 TI), but I have never thought it could improve my video workflow. I’ll look into that later.
That said, there’s something that I’m not sure about. I thought after reading the article, that the steps were:
1-Export from Premiere using MPEG2 I-frame.
2-Create AVS script.
3-Run Virtualdub to downscale video.
4-Run Encore to create DVD or use HC encoder.
It’s actually the last step the one that confuses me a little bit. At first, I thought HC was an alternative to using Encore, but then I realized that it transforms the AVI file into m2v.
So, I still need to use Encore to create the DVD ISO?
Just wanted to post an update after some testing. I followed two procedures: The one in Precomposed blog, and Jeff Bellune’s excellent tutorial on downscalling.
So far, exporting MPEG2 I-frame from Premiere, indexing with DGMPEGdec and finally saving an AVI file from Virtualdub with Avisynth gave the best results for downscalling.
From there, I had two choices: Use Encore to transcode the AVI and create the DVD ISO, or use HC Encoder to do the transcoding. This last option, gave me a transcoded m2v file that looks really good.
The only complain so far, is that I’m using permanent subtitles, and when going from AVI to m2V (already downscaled) the conversion seems to render a little bit of noise which is noticeable specially around the subs.
Unless I missed it I don’t think anybody mentioned cinema craft software which is what a lot of Hollywood uses. I found much better results exporting highest possible from premiere then using cinema craft to encode for encore. Even though the files were linked from Premier into Encore, I was able to select the cinema craft encoded files for sources and tell encore not to encode. The results were obviously much crisper text, color, and imagery.